A Collaborative Recommender System Based on Space-Time Similarities Articles uri icon

publication date

  • June 2010

start page

  • 81

end page

  • 87

issue

  • 3

volume

  • 9

international standard serial number (ISSN)

  • 1536-1268

electronic international standard serial number (EISSN)

  • 1558-2590

abstract

  • The Internet of Things (IoT) concept promises a world of networked and interconnected devices that provides relevant content to users. Recommender systems can find relevant content for users in IoT
    environments, offering a user-adapted personalized experience.
    Collaboration-based recommenders in IoT environments rely on
    user-to-object, space-time interaction patterns. This extension of that
    idea takes into account user location and interaction time to recommend
    scattered, pervasive context-embedded networked objects. The authors
    compare their proposed system to memory-based collaborative methods in
    which user similarity is based on the ratings of previously rated items.
    Their proof-of-concept implementation was used in a real-world scenario
    involving 15 students interacting with 75 objects at Carlos III
    University of Madrid.